Micro-optimization:

If we have this situation:

    jCC  L1
    jmp  L2
L1:
  ...
L2:
  ...

We can get a small performance boost by emitting this instead:

    jnCC L2
L1:
  ...
L2:
  ...

This testcase shows an example of this:

float func(float x, float y) {
    double product = (double)x * y;
    if (product == 0.0)
        return product;
    return product - 1.0;
}

llvm-svn: 101075

1 files changed, 46 insertions, 1 deletions

diff --git a/llvm/lib/Target/X86/X86InstrInfo.cpp b/llvm/lib/Target/X86/X86InstrInfo.cpp
index e67bf3d7ad2..094164ec7cf 100644
--- a/llvm/lib/Target/X86/X86InstrInfo.cpp
+++ b/llvm/lib/Target/X86/X86InstrInfo.cpp
@@ -1684,6 +1684,7 @@ bool X86InstrInfo::AnalyzeBranch(MachineBasicBlock &MBB,
   // Start from the bottom of the block and work up, examining the
   // terminator instructions.
   MachineBasicBlock::iterator I = MBB.end();
+  MachineBasicBlock::iterator UnCondBrIter = MBB.end();
   while (I != MBB.begin()) {
     --I;
     if (I->isDebugValue())
@@ -1701,6 +1702,8 @@ bool X86InstrInfo::AnalyzeBranch(MachineBasicBlock &MBB,
 
     // Handle unconditional branches.
     if (I->getOpcode() == X86::JMP_4) {
+      UnCondBrIter = I;
+
       if (!AllowModify) {
         TBB = I->getOperand(0).getMBB();
         continue;
@@ -1718,10 +1721,11 @@ bool X86InstrInfo::AnalyzeBranch(MachineBasicBlock &MBB,
         TBB = 0;
         I->eraseFromParent();
         I = MBB.end();
+        UnCondBrIter = MBB.end();
         continue;
       }
 
-      // TBB is used to indicate the unconditinal destination.
+      // TBB is used to indicate the unconditional destination.
       TBB = I->getOperand(0).getMBB();
       continue;
     }
@@ -1733,7 +1737,48 @@ bool X86InstrInfo::AnalyzeBranch(MachineBasicBlock &MBB,
 
     // Working from the bottom, handle the first conditional branch.
     if (Cond.empty()) {
+      MachineBasicBlock *TargetBB = I->getOperand(0).getMBB();
+      if (AllowModify && UnCondBrIter != MBB.end() &&
+          MBB.isLayoutSuccessor(TargetBB)) {
+        // If we can modify the code and it ends in something like:
+        //
+        //     jCC L1
+        //     jmp L2
+        //   L1:
+        //     ...
+        //   L2:
+        //
+        // Then we can change this to:
+        //
+        //     jnCC L2
+        //   L1:
+        //     ...
+        //   L2:
+        //
+        // Which is a bit more efficient.
+        // We conditionally jump to the fall-through block.
+        BranchCode = GetOppositeBranchCondition(BranchCode);
+        unsigned JNCC = GetCondBranchFromCond(BranchCode);
+        MachineBasicBlock::iterator OldInst = I;
+        --I;
+
+        BuildMI(MBB, UnCondBrIter, MBB.findDebugLoc(I), get(JNCC))
+          .addMBB(UnCondBrIter->getOperand(0).getMBB());
+        BuildMI(MBB, UnCondBrIter, MBB.findDebugLoc(I), get(X86::JMP_4))
+          .addMBB(TargetBB);
+        MBB.addSuccessor(TargetBB);
+
+        OldInst->eraseFromParent();
+        UnCondBrIter->eraseFromParent();
+
+        // Restart the analysis.
+        UnCondBrIter = MBB.end();
+        I = MBB.end();
+        continue;
+      }
+
       FBB = TBB;
+      TBB = TargetBB;
       TBB = I->getOperand(0).getMBB();
       Cond.push_back(MachineOperand::CreateImm(BranchCode));
       continue;